You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@mahout.apache.org by sr...@apache.org on 2009/03/19 16:17:08 UTC
svn commit: r756041 - in /lucene/mahout/trunk/core/src:
main/java/org/apache/mahout/clustering/dirichlet/
main/java/org/apache/mahout/clustering/dirichlet/models/
main/java/org/apache/mahout/clustering/fuzzykmeans/
main/java/org/apache/mahout/clusterin...
Author: srowen
Date: Thu Mar 19 15:17:07 2009
New Revision: 756041
URL: http://svn.apache.org/viewvc?rev=756041&view=rev
Log:
More code inspection stuff
Modified:
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletCluster.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletClusterer.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletCombiner.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletMapper.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletReducer.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletState.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/JsonDirichletStateAdapter.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/JsonModelAdapter.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/JsonModelDistributionAdapter.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/UncommonDistributions.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/AsymmetricSampledNormalDistribution.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/AsymmetricSampledNormalModel.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/Model.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/ModelDistribution.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/NormalModel.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/NormalModelDistribution.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/SampledNormalDistribution.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/SampledNormalModel.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/fuzzykmeans/FuzzyKMeansMapper.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/fuzzykmeans/FuzzyKMeansReducer.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/Cluster.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansClusterMapper.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansDriver.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansJob.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansMapper.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansReducer.java
lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansUtil.java
lucene/mahout/trunk/core/src/test/java/org/apache/mahout/clustering/dirichlet/TestDirichletClustering.java
lucene/mahout/trunk/core/src/test/java/org/apache/mahout/clustering/dirichlet/TestMapReduce.java
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletCluster.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletCluster.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletCluster.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletCluster.java Thu Mar 19 15:17:07 2009
@@ -46,7 +46,7 @@
this.totalCount += model.count();
}
- static Type typeOfModel = new TypeToken<DirichletCluster<Vector>>() {
+ static final Type typeOfModel = new TypeToken<DirichletCluster<Vector>>() {
}.getType();
public String asFormatString() {
@@ -58,7 +58,7 @@
}
@SuppressWarnings("unchecked")
- public static DirichletCluster fromFormatString(String formatString) {
+ public static DirichletCluster<Vector> fromFormatString(String formatString) {
GsonBuilder builder = new GsonBuilder();
builder.registerTypeAdapter(Vector.class, new JsonVectorAdapter());
builder.registerTypeAdapter(Model.class, new JsonModelAdapter());
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletClusterer.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletClusterer.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletClusterer.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletClusterer.java Thu Mar 19 15:17:07 2009
@@ -79,21 +79,21 @@
public class DirichletClusterer<Observation> {
// observed data
- private List<Observation> sampleData;
+ private final List<Observation> sampleData;
// the ModelDistribution for the computation
- private ModelDistribution<Observation> modelFactory;
+ private final ModelDistribution<Observation> modelFactory;
// the state of the clustering process
- private DirichletState<Observation> state;
+ private final DirichletState<Observation> state;
- private int thin;
+ private final int thin;
- private int burnin;
+ private final int burnin;
- private int numClusters;
+ private final int numClusters;
- public List<Model<Observation>[]> clusterSamples = new ArrayList<Model<Observation>[]>();
+ public final List<Model<Observation>[]> clusterSamples = new ArrayList<Model<Observation>[]>();
/**
* Create a new instance on the sample data with the given additional parameters
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletCombiner.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletCombiner.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletCombiner.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletCombiner.java Thu Mar 19 15:17:07 2009
@@ -30,6 +30,7 @@
public class DirichletCombiner extends MapReduceBase implements
Reducer<Text, Text, Text, Text> {
+ @Override
public void reduce(Text key, Iterator<Text> values,
OutputCollector<Text, Text> output, Reporter reporter) throws IOException {
}
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletMapper.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletMapper.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletMapper.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletMapper.java Thu Mar 19 15:17:07 2009
@@ -39,6 +39,7 @@
DirichletState<Vector> state;
+ @Override
public void map(WritableComparable<?> key, Text values,
OutputCollector<Text, Text> output, Reporter reporter) throws IOException {
Vector v = DenseVector.decodeFormat(values.toString());
@@ -47,7 +48,7 @@
// then pick one model by sampling a Multinomial distribution based upon them
// see: http://en.wikipedia.org/wiki/Multinomial_distribution
int k = UncommonDistributions.rMultinom(pi);
- output.collect(new Text("" + k), values);
+ output.collect(new Text(String.valueOf(k)), values);
}
public void configure(DirichletState<Vector> state) {
@@ -106,7 +107,7 @@
* @param v an Vector
* @return the Vector of probabilities
*/
- private Vector normalizedProbabilities(DirichletState<Vector> state, Vector v) {
+ private static Vector normalizedProbabilities(DirichletState<Vector> state, Vector v) {
Vector pi = new DenseVector(state.numClusters);
double max = 0;
for (int k = 0; k < state.numClusters; k++) {
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletReducer.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletReducer.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletReducer.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletReducer.java Thu Mar 19 15:17:07 2009
@@ -37,6 +37,7 @@
public Model<Vector>[] newModels;
+ @Override
public void reduce(Text key, Iterator<Text> values,
OutputCollector<Text, Text> output, Reporter reporter) throws IOException {
Integer k = new Integer(key.toString());
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletState.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletState.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletState.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/DirichletState.java Thu Mar 19 15:17:07 2009
@@ -46,7 +46,7 @@
offset = alpha_0 / numClusters;
// sample initial prior models
clusters = new ArrayList<DirichletCluster<Observation>>();
- for (Model m : modelFactory.sampleFromPrior(numClusters))
+ for (Model<?> m : modelFactory.sampleFromPrior(numClusters))
clusters.add(new DirichletCluster(m, offset));
// sample the mixture parameters from a Dirichlet distribution on the totalCounts
mixture = UncommonDistributions.rDirichlet(totalCounts());
@@ -87,8 +87,8 @@
double pdf = clusters.get(k).model.pdf(x);
double mix = mixture.get(k);
double result = mix * pdf;
- if (result < 0 || result > 1)
- System.out.print("");
+ //if (result < 0 || result > 1)
+ // System.out.print("");
return result;
}
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/JsonDirichletStateAdapter.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/JsonDirichletStateAdapter.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/JsonDirichletStateAdapter.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/JsonDirichletStateAdapter.java Thu Mar 19 15:17:07 2009
@@ -37,15 +37,15 @@
@SuppressWarnings("unchecked")
public class JsonDirichletStateAdapter implements
- JsonSerializer<DirichletState>, JsonDeserializer<DirichletState> {
+ JsonSerializer<DirichletState<?>>, JsonDeserializer<DirichletState<?>> {
- Type typeOfModel = new TypeToken<List<DirichletCluster<Vector>>>() {
+ final Type typeOfModel = new TypeToken<List<DirichletCluster<Vector>>>() {
}.getType();
- Type typeOfModelDistribution = new TypeToken<ModelDistribution<Vector>>() {
+ final Type typeOfModelDistribution = new TypeToken<ModelDistribution<Vector>>() {
}.getType();
- public JsonElement serialize(DirichletState src, Type typeOfSrc,
+ public JsonElement serialize(DirichletState<?> src, Type typeOfSrc,
JsonSerializationContext context) {
GsonBuilder builder = new GsonBuilder();
builder.registerTypeAdapter(Vector.class, new JsonVectorAdapter());
@@ -65,7 +65,7 @@
return obj;
}
- public DirichletState deserialize(JsonElement json, Type typeOfT,
+ public DirichletState<?> deserialize(JsonElement json, Type typeOfT,
JsonDeserializationContext context) throws JsonParseException {
GsonBuilder builder = new GsonBuilder();
builder.registerTypeAdapter(Vector.class, new JsonVectorAdapter());
@@ -74,7 +74,7 @@
new JsonModelDistributionAdapter());
Gson gson = builder.create();
JsonObject obj = json.getAsJsonObject();
- DirichletState state = new DirichletState();
+ DirichletState<?> state = new DirichletState();
state.numClusters = obj.get("numClusters").getAsInt();
state.offset = obj.get("offset").getAsDouble();
state.modelFactory = gson.fromJson(obj.get("modelFactory").getAsString(),
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/JsonModelAdapter.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/JsonModelAdapter.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/JsonModelAdapter.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/JsonModelAdapter.java Thu Mar 19 15:17:07 2009
@@ -20,6 +20,8 @@
import org.apache.mahout.clustering.dirichlet.models.Model;
import org.apache.mahout.matrix.Vector;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
import com.google.gson.Gson;
import com.google.gson.GsonBuilder;
@@ -33,10 +35,12 @@
import com.google.gson.JsonSerializer;
@SuppressWarnings("unchecked")
-public class JsonModelAdapter implements JsonSerializer<Model>,
- JsonDeserializer<Model> {
+public class JsonModelAdapter implements JsonSerializer<Model<?>>,
+ JsonDeserializer<Model<?>> {
- public JsonElement serialize(Model src, Type typeOfSrc,
+ private static final Logger log = LoggerFactory.getLogger(JsonModelAdapter.class);
+
+ public JsonElement serialize(Model<?> src, Type typeOfSrc,
JsonSerializationContext context) {
GsonBuilder builder = new GsonBuilder();
builder.registerTypeAdapter(Vector.class, new JsonVectorAdapter());
@@ -47,7 +51,7 @@
return obj;
}
- public Model deserialize(JsonElement json, Type typeOfT,
+ public Model<?> deserialize(JsonElement json, Type typeOfT,
JsonDeserializationContext context) throws JsonParseException {
GsonBuilder builder = new GsonBuilder();
builder.registerTypeAdapter(Vector.class, new JsonVectorAdapter());
@@ -60,8 +64,8 @@
try {
cl = ccl.loadClass(klass);
} catch (ClassNotFoundException e) {
- e.printStackTrace();
+ log.warn("Error while loading class", e);
}
- return (Model) gson.fromJson(model, cl);
+ return (Model<?>) gson.fromJson(model, cl);
}
}
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/JsonModelDistributionAdapter.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/JsonModelDistributionAdapter.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/JsonModelDistributionAdapter.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/JsonModelDistributionAdapter.java Thu Mar 19 15:17:07 2009
@@ -19,6 +19,8 @@
import java.lang.reflect.Type;
import org.apache.mahout.clustering.dirichlet.models.ModelDistribution;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
import com.google.gson.JsonDeserializationContext;
import com.google.gson.JsonDeserializer;
@@ -30,28 +32,30 @@
@SuppressWarnings("unchecked")
public class JsonModelDistributionAdapter implements
- JsonSerializer<ModelDistribution>, JsonDeserializer<ModelDistribution> {
+ JsonSerializer<ModelDistribution<?>>, JsonDeserializer<ModelDistribution<?>> {
- public JsonElement serialize(ModelDistribution src, Type typeOfSrc,
+ private static final Logger log = LoggerFactory.getLogger(JsonModelDistributionAdapter.class);
+
+ public JsonElement serialize(ModelDistribution<?> src, Type typeOfSrc,
JsonSerializationContext context) {
return new JsonPrimitive(src.getClass().getName());
}
- public ModelDistribution deserialize(JsonElement json, Type typeOfT,
+ public ModelDistribution<?> deserialize(JsonElement json, Type typeOfT,
JsonDeserializationContext context) throws JsonParseException {
ClassLoader ccl = Thread.currentThread().getContextClassLoader();
Class<?> cl = null;
try {
cl = ccl.loadClass(json.getAsString());
} catch (ClassNotFoundException e) {
- e.printStackTrace();
+ log.warn("Error while loading class", e);
}
try {
- return (ModelDistribution) cl.newInstance();
+ return (ModelDistribution<?>) cl.newInstance();
} catch (InstantiationException e) {
- e.printStackTrace();
+ log.warn("Error while creating class", e);
} catch (IllegalAccessException e) {
- e.printStackTrace();
+ log.warn("Error while creating class", e);
}
return null;
}
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/UncommonDistributions.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/UncommonDistributions.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/UncommonDistributions.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/UncommonDistributions.java Thu Mar 19 15:17:07 2009
@@ -17,8 +17,6 @@
* limitations under the License.
*/
-import java.util.ArrayList;
-import java.util.List;
import java.util.Random;
import org.apache.mahout.matrix.DenseVector;
@@ -26,12 +24,15 @@
import org.uncommons.maths.random.GaussianGenerator;
import org.uncommons.maths.random.MersenneTwisterRNG;
-public class UncommonDistributions {
+public final class UncommonDistributions {
- static final double sqrt2pi = Math.sqrt(2 * Math.PI);
+ private static final double sqrt2pi = Math.sqrt(2.0 * Math.PI);
static Random random = new MersenneTwisterRNG();
+ private UncommonDistributions() {
+ }
+
public static void init(byte[] seed) {
random = new MersenneTwisterRNG(seed);
}
@@ -52,35 +53,35 @@
double c = (k + Math.sqrt(2 * k - 1));
double lam = Math.sqrt(2 * k - 1);
double cheng = (1 + Math.log(4.5));
- double u, v, x, y, z, r;
+ double x;
do {
- u = random.nextDouble();
- v = random.nextDouble();
- y = ((1 / lam) * Math.log(v / (1 - v)));
+ double u = random.nextDouble();
+ double v = random.nextDouble();
+ double y = ((1 / lam) * Math.log(v / (1 - v)));
x = (k * Math.exp(y));
- z = (u * v * v);
- r = (b + (c * y) - x);
+ double z = (u * v * v);
+ double r = (b + (c * y) - x);
if ((r >= ((4.5 * z) - cheng)) || (r >= Math.log(z))) {
accept = true;
}
} while (!accept);
- return new Double(x / lambda);
+ return x / lambda;
} else {
//Weibull algorithm
double c = (1 / k);
double d = ((1 - k) * Math.pow(k, (k / (1 - k))));
- double u, v, z, e, x;
+ double x;
do {
- u = random.nextDouble();
- v = random.nextDouble();
- z = -Math.log(u); //generating random exponential variates
- e = -Math.log(v);
+ double u = random.nextDouble();
+ double v = random.nextDouble();
+ double z = -Math.log(u);
+ double e = -Math.log(v);
x = Math.pow(z, c);
if ((z + e) >= (d + x)) {
accept = true;
}
} while (!accept);
- return new Double(x / lambda);
+ return x / lambda;
}
}
@@ -97,8 +98,7 @@
public static double rBeta(double shape1, double shape2) {
double gam1 = rGamma(shape1, 1);
double gam2 = rGamma(shape2, 1);
- double result = gam1 / (gam1 + gam2);
- return result;
+ return gam1 / (gam1 + gam2);
}
@@ -112,9 +112,9 @@
* @return a Vector of samples
*/
public static Vector rBeta(int K, double shape1, double shape2) {
- List<Double> params = new ArrayList<Double>(2);
- params.add(shape1);
- params.add(Math.max(0, shape2));
+ //List<Double> params = new ArrayList<Double>(2);
+ //params.add(shape1);
+ //params.add(Math.max(0, shape2));
Vector result = new DenseVector(K);
for (int i = 0; i < K; i++)
result.set(i, rBeta(shape1, shape2));
@@ -128,7 +128,7 @@
* @return a double sample
*/
public static double rChisq(double df) {
- double result = 0;
+ double result = 0.0;
for (int i = 0; i < df; i++) {
double sample = rNorm(0, 1);
result += sample * sample;
@@ -164,8 +164,7 @@
double xms = (x - m) / s;
double ex = (xms * xms) / 2;
double exp = Math.exp(-ex);
- double result = exp / (sqrt2pi * s);
- return result;
+ return exp / (sqrt2pi * s);
}
/**
@@ -232,10 +231,9 @@
double q = -Math.log(1 - p);
double sum = 0;
int x = 0;
- double u, e;
while (sum <= q) {
- u = random.nextDouble();
- e = -Math.log(u); //exponential random variate
+ double u = random.nextDouble();
+ double e = -Math.log(u);
sum += (e / (n - x));
x += 1;
}
@@ -253,11 +251,10 @@
*/
public static Vector rDirichlet(Vector alpha) {
Vector r = alpha.like();
- double a;
double total = alpha.zSum();
double remainder = 1;
for (int i = 0; i < r.size(); i++) {
- a = alpha.get(i);
+ double a = alpha.get(i);
total -= a;
double beta = rBeta(a, Math.max(0, total));
double p = beta * remainder;
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/AsymmetricSampledNormalDistribution.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/AsymmetricSampledNormalDistribution.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/AsymmetricSampledNormalDistribution.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/AsymmetricSampledNormalDistribution.java Thu Mar 19 15:17:07 2009
@@ -30,10 +30,7 @@
public class AsymmetricSampledNormalDistribution implements
ModelDistribution<Vector> {
-
- /* (non-Javadoc)
- * @see org.apache.mahout.clustering.dirichlet.ModelDistribution#sampleFromPrior(int)
- */
+ @Override
public Model<Vector>[] sampleFromPrior(int howMany) {
Model<Vector>[] result = new AsymmetricSampledNormalModel[howMany];
for (int i = 0; i < howMany; i++) {
@@ -48,9 +45,7 @@
return result;
}
- /* (non-Javadoc)
- * @see org.apache.mahout.clustering.dirichlet.ModelDistribution#sampleFromPosterior(org.apache.mahout.clustering.dirichlet.Model<Observation>[])
- */
+ @Override
public Model<Vector>[] sampleFromPosterior(Model<Vector>[] posterior) {
Model<Vector>[] result = new AsymmetricSampledNormalModel[posterior.length];
for (int i = 0; i < posterior.length; i++) {
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/AsymmetricSampledNormalModel.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/AsymmetricSampledNormalModel.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/AsymmetricSampledNormalModel.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/AsymmetricSampledNormalModel.java Thu Mar 19 15:17:07 2009
@@ -19,6 +19,9 @@
import org.apache.mahout.matrix.Vector;
public class AsymmetricSampledNormalModel implements Model<Vector> {
+
+ private static final double sqrt2pi = Math.sqrt(2.0 * Math.PI);
+
// the parameters
public Vector mean;
@@ -52,9 +55,7 @@
return new AsymmetricSampledNormalModel(mean, sd);
}
- /* (non-Javadoc)
- * @see org.apache.mahout.clustering.dirichlet.Model#observe(java.lang.Object)
- */
+ @Override
public void observe(Vector x) {
s0++;
if (s1 == null)
@@ -67,9 +68,7 @@
s2 = s2.plus(x.times(x));
}
- /* (non-Javadoc)
- * @see org.apache.mahout.clustering.dirichlet.Model#computeParameters()
- */
+ @Override
public void computeParameters() {
if (s0 == 0)
return;
@@ -97,30 +96,26 @@
double sd2 = sd * sd;
double exp = -(x.dot(x) - 2 * x.dot(mean) + mean.dot(mean)) / (2 * sd2);
double ex = Math.exp(exp);
- double pdf = ex / (sd * Math.sqrt(2 * Math.PI));
- return pdf;
+ return ex / (sd * sqrt2pi);
}
- /* (non-Javadoc)
- * @see org.apache.mahout.clustering.dirichlet.Model#pdf(java.lang.Object)
- */
+ @Override
public double pdf(Vector x) {
// return the product of the two component pdfs
assert x.size() == 2;
double pdf0 = pdf(x, sd.get(0));
double pdf1 = pdf(x, sd.get(1));
- if (pdf0 < 0 || pdf0 > 1 || pdf1 < 0 || pdf1 > 1)
- System.out.print("");
+ //if (pdf0 < 0 || pdf0 > 1 || pdf1 < 0 || pdf1 > 1)
+ // System.out.print("");
return pdf0 * pdf1;
}
- /* (non-Javadoc)
- * @see org.apache.mahout.clustering.dirichlet.Model#count()
- */
+ @Override
public int count() {
return s0;
}
+ @Override
public String toString() {
StringBuilder buf = new StringBuilder();
buf.append("asnm{n=").append(s0).append(" m=[");
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/Model.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/Model.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/Model.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/Model.java Thu Mar 19 15:17:07 2009
@@ -28,13 +28,13 @@
*
* @param x an Observation from the posterior
*/
- public abstract void observe(Observation x);
+ void observe(Observation x);
/**
* Compute a new set of posterior parameters based upon the Observations
* that have been observed since my creation
*/
- public abstract void computeParameters();
+ void computeParameters();
/**
* Return the probability that the observation is described by this model
@@ -42,12 +42,12 @@
* @param x an Observation from the posterior
* @return the probability that x is in the receiver
*/
- public abstract double pdf(Observation x);
+ double pdf(Observation x);
/**
* Return the number of observations that have been observed by this model
*
* @return an int
*/
- public abstract int count();
+ int count();
}
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/ModelDistribution.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/ModelDistribution.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/ModelDistribution.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/ModelDistribution.java Thu Mar 19 15:17:07 2009
@@ -21,7 +21,7 @@
/**
* A model distribution allows us to sample a model from its prior distribution.
*/
-public interface ModelDistribution<Observation> {
+public interface ModelDistribution<O> {
/**
* Return a list of models sampled from the prior
@@ -29,7 +29,7 @@
* @param howMany the int number of models to return
* @return a Model<Observation>[] representing what is known apriori
*/
- public Model<Observation>[] sampleFromPrior(int howMany);
+ Model<O>[] sampleFromPrior(int howMany);
/**
* Return a list of models sampled from the posterior
@@ -37,6 +37,6 @@
* @param posterior the Model<Observation>[] after observations
* @return a Model<Observation>[] representing what is known apriori
*/
- public Model<Observation>[] sampleFromPosterior(Model<Observation>[] posterior);
+ Model<O>[] sampleFromPosterior(Model<O>[] posterior);
}
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/NormalModel.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/NormalModel.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/NormalModel.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/NormalModel.java Thu Mar 19 15:17:07 2009
@@ -20,6 +20,8 @@
public class NormalModel implements Model<Vector> {
+ private static final double sqrt2pi = Math.sqrt(2.0 * Math.PI);
+
// the parameters
public Vector mean;
@@ -27,17 +29,13 @@
// the observation statistics, initialized by the first observation
int s0 = 0;
-
Vector s1;
-
Vector s2;
public NormalModel() {
- super();
}
public NormalModel(Vector mean, double sd) {
- super();
this.mean = mean;
this.sd = sd;
this.s0 = 0;
@@ -53,9 +51,7 @@
return new NormalModel(mean, sd);
}
- /* (non-Javadoc)
- * @see org.apache.mahout.clustering.dirichlet.Model#observe(java.lang.Object)
- */
+ @Override
public void observe(Vector x) {
s0++;
if (s1 == null)
@@ -68,9 +64,7 @@
s2 = s2.plus(x.times(x));
}
- /* (non-Javadoc)
- * @see org.apache.mahout.clustering.dirichlet.Model#computeParameters()
- */
+ @Override
public void computeParameters() {
if (s0 == 0)
return;
@@ -82,31 +76,28 @@
sd = Double.MIN_VALUE;
}
- /* (non-Javadoc)
- * @see org.apache.mahout.clustering.dirichlet.Model#pdf(java.lang.Object)
- */
+ @Override
public double pdf(Vector x) {
assert x.size() == 2;
double sd2 = sd * sd;
double exp = -(x.dot(x) - 2 * x.dot(mean) + mean.dot(mean)) / (2 * sd2);
double ex = Math.exp(exp);
- return ex / (sd * Math.sqrt(2 * Math.PI));
+ return ex / (sd * sqrt2pi);
}
- /* (non-Javadoc)
- * @see org.apache.mahout.clustering.dirichlet.Model#count()
- */
+ @Override
public int count() {
return s0;
}
+ @Override
public String toString() {
StringBuilder buf = new StringBuilder();
buf.append("nm{n=").append(s0).append(" m=[");
if (mean != null)
for (int i = 0; i < mean.cardinality(); i++)
buf.append(String.format("%.2f", mean.get(i))).append(", ");
- buf.append("] sd=").append(String.format("%.2f", sd)).append("}");
+ buf.append("] sd=").append(String.format("%.2f", sd)).append('}');
return buf.toString();
}
}
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/NormalModelDistribution.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/NormalModelDistribution.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/NormalModelDistribution.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/NormalModelDistribution.java Thu Mar 19 15:17:07 2009
@@ -26,14 +26,7 @@
*/
public class NormalModelDistribution implements ModelDistribution<Vector> {
- public NormalModelDistribution() {
- super();
- }
-
-
- /* (non-Javadoc)
- * @see org.apache.mahout.clustering.dirichlet.ModelDistribution#sampleFromPrior(int)
- */
+ @Override
public Model<Vector>[] sampleFromPrior(int howMany) {
Model<Vector>[] result = new NormalModel[howMany];
for (int i = 0; i < howMany; i++)
@@ -41,9 +34,7 @@
return result;
}
- /* (non-Javadoc)
- * @see org.apache.mahout.clustering.dirichlet.ModelDistribution#sampleFromPosterior(org.apache.mahout.clustering.dirichlet.Model<Observation>[])
- */
+ @Override
public Model<Vector>[] sampleFromPosterior(Model<Vector>[] posterior) {
Model<Vector>[] result = new NormalModel[posterior.length];
for (int i = 0; i < posterior.length; i++) {
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/SampledNormalDistribution.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/SampledNormalDistribution.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/SampledNormalDistribution.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/SampledNormalDistribution.java Thu Mar 19 15:17:07 2009
@@ -26,12 +26,9 @@
* DirichletCluster algorithm. Uses a Normal Distribution to sample the prior
* model values.
*/
-public class SampledNormalDistribution extends NormalModelDistribution
- implements ModelDistribution<Vector> {
+public class SampledNormalDistribution extends NormalModelDistribution {
- /* (non-Javadoc)
- * @see org.apache.mahout.clustering.dirichlet.ModelDistribution#sampleFromPrior(int)
- */
+ @Override
public Model<Vector>[] sampleFromPrior(int howMany) {
Model<Vector>[] result = new SampledNormalModel[howMany];
for (int i = 0; i < howMany; i++) {
@@ -43,9 +40,7 @@
return result;
}
- /* (non-Javadoc)
- * @see org.apache.mahout.clustering.dirichlet.ModelDistribution#sampleFromPosterior(org.apache.mahout.clustering.dirichlet.Model<Observation>[])
- */
+ @Override
public Model<Vector>[] sampleFromPosterior(Model<Vector>[] posterior) {
Model<Vector>[] result = new SampledNormalModel[posterior.length];
for (int i = 0; i < posterior.length; i++) {
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/SampledNormalModel.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/SampledNormalModel.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/SampledNormalModel.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/dirichlet/models/SampledNormalModel.java Thu Mar 19 15:17:07 2009
@@ -18,7 +18,7 @@
import org.apache.mahout.matrix.Vector;
-public class SampledNormalModel extends NormalModel implements Model<Vector> {
+public class SampledNormalModel extends NormalModel {
public SampledNormalModel() {
super();
@@ -28,13 +28,14 @@
super(mean, sd);
}
+ @Override
public String toString() {
StringBuilder buf = new StringBuilder();
buf.append("snm{n=").append(s0).append(" m=[");
if (mean != null)
for (int i = 0; i < mean.cardinality(); i++)
buf.append(String.format("%.2f", mean.get(i))).append(", ");
- buf.append("] sd=").append(String.format("%.2f", sd)).append("}");
+ buf.append("] sd=").append(String.format("%.2f", sd)).append('}');
return buf.toString();
}
@@ -42,6 +43,7 @@
* Return an instance with the same parameters
* @return an SampledNormalModel
*/
+ @Override
NormalModel sample() {
return new SampledNormalModel(mean, sd);
}
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/fuzzykmeans/FuzzyKMeansMapper.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/fuzzykmeans/FuzzyKMeansMapper.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/fuzzykmeans/FuzzyKMeansMapper.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/fuzzykmeans/FuzzyKMeansMapper.java Thu Mar 19 15:17:07 2009
@@ -36,8 +36,7 @@
public class FuzzyKMeansMapper extends MapReduceBase implements
Mapper<WritableComparable<?>, Text, Text, Text> {
- private static final Logger log = LoggerFactory
- .getLogger(FuzzyKMeansMapper.class);
+ private static final Logger log = LoggerFactory.getLogger(FuzzyKMeansMapper.class);
protected List<SoftCluster> clusters;
@@ -57,11 +56,6 @@
this.clusters = clusters;
}
- /*
- * (non-Javadoc)
- *
- * @see org.apache.hadoop.mapred.MapReduceBase#configure(org.apache.hadoop.mapred.JobConf)
- */
@Override
public void configure(JobConf job) {
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/fuzzykmeans/FuzzyKMeansReducer.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/fuzzykmeans/FuzzyKMeansReducer.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/fuzzykmeans/FuzzyKMeansReducer.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/fuzzykmeans/FuzzyKMeansReducer.java Thu Mar 19 15:17:07 2009
@@ -32,15 +32,10 @@
import org.apache.hadoop.mapred.Reporter;
import org.apache.mahout.matrix.AbstractVector;
import org.apache.mahout.matrix.Vector;
-import org.slf4j.Logger;
-import org.slf4j.LoggerFactory;
public class FuzzyKMeansReducer extends MapReduceBase implements
Reducer<Text, Text, Text, Text> {
- private static final Logger log = LoggerFactory
- .getLogger(FuzzyKMeansReducer.class);
-
protected Map<String, SoftCluster> clusterMap;
@Override
@@ -79,11 +74,6 @@
.formatCluster(cluster)));
}
- /*
- * (non-Javadoc)
- *
- * @see org.apache.hadoop.mapred.MapReduceBase#configure(org.apache.hadoop.mapred.JobConf)
- */
@Override
public void configure(JobConf job) {
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/Cluster.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/Cluster.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/Cluster.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/Cluster.java Thu Mar 19 15:17:07 2009
@@ -99,7 +99,7 @@
*/
public static void configure(JobConf job) {
try {
- final ClassLoader ccl = Thread.currentThread().getContextClassLoader();
+ ClassLoader ccl = Thread.currentThread().getContextClassLoader();
Class<?> cl = ccl.loadClass(job.get(DISTANCE_MEASURE_KEY));
measure = (DistanceMeasure) cl.newInstance();
measure.configure(job);
@@ -213,8 +213,6 @@
/**
* Construct a new clsuter with the given id as identifier
- *
- * @param identifier
*/
public Cluster(String clusterId) {
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansClusterMapper.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansClusterMapper.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansClusterMapper.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansClusterMapper.java Thu Mar 19 15:17:07 2009
@@ -26,6 +26,7 @@
import org.apache.mahout.matrix.Vector;
public class KMeansClusterMapper extends KMeansMapper {
+ @Override
public void map(WritableComparable<?> key, Text values,
OutputCollector<Text, Text> output, Reporter reporter) throws IOException {
Vector point = AbstractVector.decodeVector(values.toString());
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansDriver.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansDriver.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansDriver.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansDriver.java Thu Mar 19 15:17:07 2009
@@ -55,7 +55,7 @@
/**
* Run the job using supplied arguments
- *
+ *
* @param input the directory pathname for input points
* @param clustersIn the directory pathname for initial & computed clusters
* @param output the directory pathname for output points
@@ -88,7 +88,7 @@
/**
* Run the job using supplied arguments
- *
+ *
* @param input the directory pathname for input points
* @param clustersIn the directory pathname for iniput clusters
* @param clustersOut the directory pathname for output clusters
@@ -135,7 +135,7 @@
/**
* Run the job using supplied arguments
- *
+ *
* @param input the directory pathname for input points
* @param clustersIn the directory pathname for input clusters
* @param output the directory pathname for output points
@@ -193,4 +193,3 @@
return converged;
}
}
-
\ No newline at end of file
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansJob.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansJob.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansJob.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansJob.java Thu Mar 19 15:17:07 2009
@@ -30,15 +30,15 @@
}
public static void main(String[] args) throws IOException {
- int index = 0;
if (args.length != 7) {
- System.out.println("Expected number of arguments 10 and received:"
+ System.err.println("Expected number of arguments 10 and received:"
+ args.length);
- System.out
+ System.err
.println("Usage:input clustersIn output measureClass convergenceDelta maxIterations numCentroids");
- System.exit(1);
+ throw new IllegalArgumentException();
}
+ int index = 0;
String input = args[index++];
String clusters = args[index++];
String output = args[index++];
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansMapper.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansMapper.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansMapper.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansMapper.java Thu Mar 19 15:17:07 2009
@@ -51,11 +51,6 @@
this.clusters = clusters;
}
- /*
- * (non-Javadoc)
- *
- * @see org.apache.hadoop.mapred.MapReduceBase#configure(org.apache.hadoop.mapred.JobConf)
- */
@Override
public void configure(JobConf job) {
super.configure(job);
@@ -66,7 +61,7 @@
KMeansUtil.configureWithClusterInfo(job.get(Cluster.CLUSTER_PATH_KEY),
clusters);
- if (clusters.size() == 0)
+ if (clusters.isEmpty())
throw new NullPointerException("Cluster is empty!!!");
}
}
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansReducer.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansReducer.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansReducer.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansReducer.java Thu Mar 19 15:17:07 2009
@@ -53,11 +53,6 @@
.formatCluster(cluster)));
}
- /*
- * (non-Javadoc)
- *
- * @see org.apache.hadoop.mapred.MapReduceBase#configure(org.apache.hadoop.mapred.JobConf)
- */
@Override
public void configure(JobConf job) {
@@ -70,7 +65,7 @@
clusters);
setClusterMap(clusters);
- if (clusterMap.size() == 0)
+ if (clusterMap.isEmpty())
throw new NullPointerException("Cluster is empty!!!");
}
@@ -80,7 +75,6 @@
clusterMap.put(cluster.getIdentifier(), cluster);
}
clusters.clear();
- clusters = null;
}
public void config(List<Cluster> clusters) {
Modified: lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansUtil.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansUtil.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansUtil.java (original)
+++ lucene/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/kmeans/KMeansUtil.java Thu Mar 19 15:17:07 2009
@@ -31,14 +31,15 @@
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
-public class KMeansUtil {
+public final class KMeansUtil {
+
private static final Logger log = LoggerFactory.getLogger(KMeansUtil.class);
+ private KMeansUtil() {
+ }
+
/**
* Configure the mapper with the cluster info
- *
- * @param job
- * @param clusters
*/
public static void configureWithClusterInfo(String clusterPathStr,
List<Cluster> clusters) {
@@ -50,6 +51,7 @@
// filter out the files
PathFilter clusterFileFilter = new PathFilter() {
+ @Override
public boolean accept(Path path) {
return path.getName().startsWith("part");
}
@@ -88,7 +90,6 @@
} catch (IOException e) {
log.info("Exception occurred in loading clusters:", e);
- e.printStackTrace();
throw new RuntimeException(e);
}
}
Modified: lucene/mahout/trunk/core/src/test/java/org/apache/mahout/clustering/dirichlet/TestDirichletClustering.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/test/java/org/apache/mahout/clustering/dirichlet/TestDirichletClustering.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/test/java/org/apache/mahout/clustering/dirichlet/TestDirichletClustering.java (original)
+++ lucene/mahout/trunk/core/src/test/java/org/apache/mahout/clustering/dirichlet/TestDirichletClustering.java Thu Mar 19 15:17:07 2009
@@ -33,11 +33,6 @@
private List<Vector> sampleData;
- /*
- * (non-Javadoc)
- *
- * @see junit.framework.TestCase#setUp()
- */
@Override
protected void setUp() throws Exception {
super.setUp();
Modified: lucene/mahout/trunk/core/src/test/java/org/apache/mahout/clustering/dirichlet/TestMapReduce.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/core/src/test/java/org/apache/mahout/clustering/dirichlet/TestMapReduce.java?rev=756041&r1=756040&r2=756041&view=diff
==============================================================================
--- lucene/mahout/trunk/core/src/test/java/org/apache/mahout/clustering/dirichlet/TestMapReduce.java (original)
+++ lucene/mahout/trunk/core/src/test/java/org/apache/mahout/clustering/dirichlet/TestMapReduce.java Thu Mar 19 15:17:07 2009
@@ -78,9 +78,6 @@
output.close();
}
- /* (non-Javadoc)
- * @see junit.framework.TestCase#setUp()
- */
@Override
protected void setUp() throws Exception {
super.setUp();